LIDAR Line Detection Using Time-Offset Segments Against Cross-Talk
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing LIDAR systems suffer from optical cross-talk between detection channels, leading to inaccurate 3D point cloud representations due to false signals from high-reflectivity objects, which complicates and increases the cost of these systems.
Innovation Solution
A method for operating a LIDAR unit that emits segments of electromagnetic radiation temporally offset along a line, allowing for the differentiation between real signals and cross-talk signals by comparing signal strengths with a threshold value, using sequences with varying temporal distances between segments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If a line detector is used to detect the entire line simultaneously, then the detection speed is improved, but optical cross-talk between adjacent detection channels occurs causing false signal magnification
Solution Approach 1:
The electromagnetic radiation line is segmented into multiple temporally offset segments that are emitted sequentially rather than simultaneously. Each segment is detected by the line detector at different time points, preventing optical cross-talk between adjacent detection channels while maintaining the speed advantage of line detection.
Solution Approach 2:
The line detector is activated in periodic pulses corresponding to each emitted radiation segment. By synchronizing the detector activation with the periodic emission of segmented radiation, the system achieves fast detection while eliminating cross-talk through temporal separation of detection events.
2Reliability
If multiple emission sequences with different temporal distances are used to identify real signals, then the reliability of signal identification is improved, but the measurement time is increased
Solution Approach 1:
Multiple emission sequences with different temporal distances are prepared in advance, with their temporal parameters pre-configured. This preliminary setup allows the system to quickly switch between different sequences during operation without requiring real-time calculation or adjustment, reducing the time penalty for improved reliability.
Solution Approach 2:
The system dynamically selects which emission sequence to use based on real-time detection needs and environmental conditions. By making the temporal distance between segments adjustable rather than fixed, the system can optimize the balance between reliability and measurement time for different operational scenarios.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach effectively distinguishes between real and cross-talk signals, reducing artifacts in 3D point clouds and enhancing the accuracy and cost-effectiveness of LIDAR systems, particularly in autonomous vehicles.
Implementation Method 1
an emitter unit (2) which is designed to emit electromagnetic radiation (4)
Implementation Method 2
first signals (10), which are based on emitted electromagnetic radiation (4) of segments (5) of the first sequence (16), reflected at objects (100) and striking the detector (3)
Data Source
AI summary
An emitter unit is designed to emit segments of electromagnetic radiation temporally offset to one another along a line. A detector includes a plurality of linearly situated detector channels. A first sequence and at least one second sequence of segments are emitted. The sequences differ at least with respect to a temporal distance between two successive segments. First signals and second signals are detected on the basis of segments of the first and second sequence reflected at objects and striking the detector. Signal strengths of first and second signals are added together in order to obtain sum signals for each detection channel. Signal strengths of the sum signals are compared with a predefinable threshold value and identified as real if they are greater than the predefinable threshold value.